Human motion analysis is an important tool for assessing movement, rehabilitation progress, fall risk, progression of neurodegenerative diseases, and classifying gait patterns. Advancements in artificial intelligence models and high-performance computing technologies have given rise to marker-less human motion analysis that determine point correspondences between an array of cameras and estimate 3D joint coordinates using triangulation. However, existing methods have not considered the physical setup and design of a marker-less human motion analysis tool that could be deployed in an institutional environment for active use, such as an institutional hallway where individuals pass regularly on a daily basis (i.e., Smart Hallway). In this thesis, camera locations were modelled, four machine vision grade cameras connected to an NVIDIA Jetson AGX were set up in a simulated institutional hallway environment, and custom software was written to capture synchronized 60 frame per second video of a participant walking through the Smart Hallway capture volume. Software was also written to calculate 3D joint coordinates and extract outcome measures for various test conditions. These outcome measures were compared to ground truth body segment length measurements obtained from direct measurement and ground truth foot event timings obtained from direct observation. Body segment length measurements were within 1.56 (SD=2.77) cm of ground truth values. AI-based stride parameters were comparable with ground truth foot event timings and the implemented foot event detection algorithm was within 4 frames (67 ms), with an absolute error of 3 frames (50 ms) on the ground truth foot event labels. The Smart Hallway can be deployed in an unobtrusive manner and be temporally and spatially calibrated with ease. This multi-camera marker-less approach is viable for calculating useful outcome measures for clinical decision making. With these findings, marker-less motion capture is viable for non-invasive human motion analysis and compares well with marker-based systems. With future research and innovations, marker-less motion analysis will be the ideal approach for human motion analysis that requires minimal human resource to collect meaningful information.

Background: In recent years, research in the field of Augmented Reality (AR) in cultural heritage has been rapidly expanding, due to the advancement of technology and availability of cheaper “off the shelf” hardware. It is, amongst other things, being used as a means to increase availability and regain the public’s interest in cultural heritage.Objectives: This study compares marker-based and marker-less AR in perceived usability and perceived performance through a user study. Methods: With the use of the software Unity3D and Vuforia, two AR applications were implemented. Both applications display a model of an 18th-century Swedish warship, based on a wooden ship model, each using one of the two AR methods. The digital model was remade in Autodesk Maya, to suit the needs of an AR application used on mobile devices. The applications were evaluated in a user study with 14 participants. Each participant was asked to perform a simple task of walking around the displayed ship and then answering a questionnaire on usability. This process was done for both applications, followed by a post-experiment questionnaire on perceived performance where the two methods were compared. Results: The result of the study showed that both applications were perceived as usable and well performing. The result of the usability questionnaire showed that the applications were considered usable, with an average of 90.5 points for marker-based AR and 86.8 points for marker-less AR on a 0-100 point scale. Regarding performance, the marker-based method was perceived as better performing. Conclusions: The participants felt that with just a few instructions, the applications were easy to use, even though 50% of them had no previous experience in using AR, that it could enhance a museum exhibition. Possible further development of the app would be to complete the ship-model by adding more details that are currently missing.

Human gait analysis is an essential indicator for physical and neuroglial health of an individual. Recent developments in deep-learning approaches to computer vision make possible new techniques for body segment and joint detection from photos and video frames. In this thesis, we propose a deep learning approach for non-invasive video-based gait analysis using two RGB cameras that would be suitable for routine gait monitoring in senior care and rehabilitation centers. Due to modularity and the low cost of implementation, it is considered an affordable solution for such centers. Furthermore, since the solution does not require any markers or sensors to be worn, it is a pervasive and easy method for daily usage. Our proposed deep-learning approach starts by calibrating both the intrinsic and extrinsic parameters of the cameras. Next, video streams captured from two RGB cameras are used as input, and OpenPose and HyperPose deep-learning frameworks are used to localize the main body key points, including the joints and skeleton based on Body 25 and COCO models, respectively. The 2D parameter outputs from the frameworks are triangulated into 3D vector spaces for further analysis. In order to reduce the noises in our data, we applied median and dual pass butter worth filters to the data. Finally gait parameters has been extracted measured and compared to the manually evaluated ground truth data which has been capture via manual measurement of a domain expert. The approach was evaluated in a laboratory setting similar to an institutional hallway in five types of trials: walking back and forth in a straight line while turning out of frame, walking back and forth in a straight line while turning in frame, circular walking, walking with a cane and a walker. The method brings promising results compared to more expensive and restrictive approaches that use up to 16 cameras and require markers or sensors.

Single view reconstruction is fundamentally an under-constrained problem. We aim to develop new approaches to model human face and motion with model priors that restrict the space of possible solutions. First, we develop a novel approach to recover the 3D shape from a single view image under challenging conditions, such as large variations in illumination and pose. The problem is addressed by employing the techniques of non-linear manifold embedding and alignment. Specifically, the local image models for each patch of facial images and the local surface models for each patch of 3D shape are learned using a non-linear dimensionality reduction technique, and the correspondences between these local models are then learned by a manifold alignment method. Local models successfully remove the dependency of large training databases for human face modeling. By combining the local shapes, the global shape of a face can be reconstructed directly from a single linear system of equations via least square. Unfortunately, this learning-based approach cannot be successfully applied to the problem of human motion modeling due to the internal and external variations in single view video-based marker-less motion capture. Therefore, we introduce a new model-based approach for capturing human motion using a stream of depth images from a single depth sensor. While a depth sensor provides metric 3D information, using a single sensor, instead of a camera array, results in a view-dependent and incomplete measurement of object motion. We develop a novel two-stage template fitting algorithm that is invariant to subject size and view-point variations, and robust to occlusions. Starting from a known pose, our algorithm first estimates a body configuration through temporal registration, which is used to search the template motion database for a best match. The best match body configuration as well as its corresponding surface mesh model are deformed to fit the input depth map, filling in the part that is occluded from the input and compensating for differences in pose and body-size between the input image and the template. Our approach does not require any makers, user-interaction, or appearance-based tracking. Experiments show that our approaches can achieve good modeling results for human face and motion, and are capable of dealing with variety of challenges in single view reconstruction, e.g., occlusion.

Augmented reality (AR) has revolutionized the way people view the real world, AR has been used across a range of sectors. Recently, researchers examined the possibilities for improving user experience with augmented reality. However, there are few studies on adoption of AR users' interactions with online data resources. The aim of this study is proposing a mobile augmented reality interface for users to interact and engage with online data. The prototype is based on the framework of a PEAR (Public Engagement Augmented Reality) initiative for further AR development. PEAR framework provides an AR extension and enables users to engage with online information through AR representation [1]. This prototype was developed and implemented using Unity game engines C# and Vuforia SDK on the front-end, both NodeJS servers with MongoDB were used on the back-end. The prototype was tested and then used in a 2-week user study to analyse and validate the framework.

L’objectif du travail de cette thèse est la contribution au développement de nouvelles techniques dans le domaine dessystèmes de détection sans marqueur pour une utilisation dans trois domaines vitaux de la santé en utilisant des capteursinnovants et peu coûteux. Pour la réalisation de nos objectifs nous avons eu recours principalement à de l’électroniqueembarquées et du traitement du signal en utilisant le capteur Kinect. Des résultats encourageants ont été obtenus et sontprésentés tout au long de cette thèse. Dans la première partie de ce travail, nous présentons un nouveau système desurveillance visuelle sans marqueur en temps réel pour détecter et suivre les personnes âgées et surveiller leurs activitésdans leur environnement intérieur en utilisant un réseau de capteurs Kinect. Le système identifie également l’événementde chute des personnes âgées sous surveillance. Dans la deuxième partie nous utilisons également le capteur Kinectmais cette fois ci pour la détection sans marqueur des mouvements de la tête d’un patient lors d’un examen utilisant LaTomographie par Emission de Positons (CT/PET) du cerveau. Ce travail est basé sur la compensation de la dégradationde l’image TEP due aux mouvements de la tête du patient. Pour nos essais un cobaye dit « fantôme » a été réalisé,les résultats sur le fantôme sont prometteur ce qui a donné lieu à un test sur un vrai patient volontaire. Les résultatsfinaux montrent l’efficacité de ce nouveau système. La troisième partie du travail présente la mise en oeuvre d’un nouveausystème intelligent pour contrôler un fauteuil roulant électrique par des mouvements spéciaux de la tête toujours sansmarqueur. Un algorithme adapté est conçu pour détecter en continu les degrés des mouvements du visage en utilisant lecapteur Kinect. Fautes de fauteuil roulant électrique, le système a été testé sur un véhicule radio commandé. / The objective of the work of this thesis is the contribution in developing novel technical methods in the field of marker-lesssensing systems for use in three vital health areas by using new inexpensive sensors. Several scientific areas are involvedin achieving our objective such as; electronics and signal processing by using the Kinect sensor. Encouraging results wereachieved as presented throughout this thesis. In the first part of this work we present a new real-time marker-less visualsurveillance system for detecting and tracking seniors and monitoring their activities in the indoor environment by usingnetwork of Kinect sensors. The system also identifies the fall event with the elderly. In the second part, we present anew approach for a marker-less movement detection system for influential head movements in the brain Positron EmissionTomography imaging (CT/PET) by employing the Kinect sensor. This work addresses the compensation of the PET imagedegradation due to subject’s head movements. A developed particular phantom and volunteer studies were carried out.The experimental results show the effectiveness of this new system. The third part of the work presents the design andimplementation of a new smart system for controlling an electric wheelchair by special mark-less head movements. Anadaptable algorithm is designed to continuously detect the rotation degrees of the face pose using the Kinect sensor inreal-time that are interpreted as controlling signals through a hardware interface for the electric wheelchair actuators.